Recombinant DNA technologies (also known as genetic, protein and metabolic engineering) allow the production of a wide range of peptides and proteins in cells which do not naturally produce such peptides and proteins. The introduction of a heterologous as well as of a homologous gene, along with controlling sequences, in the selected host could lead to large accumulation of useful products: for example, proteins, enzymes, hormones or antigens. The availability of significant amounts of proteins is often a highly desirable goal. For example, in the pharmaceutical field, pre-clinical and clinical trials often require substantial amounts of potentially interesting recombinant proteins. Some recombinant products are already available on the market (such as growth hormone, Tissue Plasminogen Activator, hepatitis B virus vaccine, interferons, and erythropoietin), and many more are currently in the last phase of clinical trials. Production of recombinant proteins also has applications in other industrial sectors. Some recombinant products are used in the food industry (e.g., β-galactosidase, chymosin, amylases, glucoamylase, and amyloglucosidase) as well as in the textile and paper industries (e.g., proteases, amylases, cellulases, lipases, and catalases). Recombinant enzymes are useful as detergents (proteases, lipases and surfactants), and their characteristics of stereo-specificity are exploited in a wide number of bioconversions, often yielding desired chiral compounds. A promising field is the application of recombinant enzymes for the development of biosensors. The potential applications of the biosensor technology range from human health to environmental monitoring and the control of industrial bioprocesses. Finally, a very interesting class of heterologous expressed genes are those which give new metabolic abilities to the host cells, allowing the use of non-conventional substrates (e.g., whey, phenols, starch, lignin, and cellulose). On the other hand, metabolic engineering could increase the production of fine chemical metabolites, such as organic acids (e.g., lactic acid), amino acids (e.g., glutamic acid), vitamins, and solvents (e.g., ethanol, 1-2 propanediol, and butanol).
The first requisite for a successful process based on engineered cells for the production of recombinant gene products concerns the choice of the host and of the expression vector. The choice must consider different factors such as product complexity, host characteristics and production level of the desired protein. From a chronological point of view, the first hosts used for the production of heterologous proteins were prokaryotes: Escherichia coli and Bacillus subtilis; later on, eukaryotic host cells were also used, particularly Saccharomyces cerevisiae (S. cerevisiae).
The yeast S. cerevisiae, commonly considered a safe organism, has been used for centuries in food processes. Moreover, it is a well-known microorganism: its genome has been completely sequenced and its physiology and biochemistry have been studied for a long time. This yeast is able to perform some post-translational modifications of heterologous protein products, which often are important for retaining biological activity; such posttranslational modifications cannot usually be obtained using a prokaryotic host. Finally, it is possible to drive the secretion of the desired product directly into the growth medium, thus improving the large-scale recovery and purification of a correctly folded, homogeneous product. Along the years, S. cerevisiae has been developed as a host for the production of both heterologous and homologous gene products with applications in many important fields of modern society (e.g., health care, pharmaceuticals, environment, agriculture, food, and chemistry) (reviewed by: Romanos, M. A., Scorer, C. A. and Clare, J. J. (1992) Yeast 8:423-488; Sudbery, P. (1996) Curr. Op. Biotechnol. 7:517-524; and Lin Cereghino, G. P. and Cregg, J. P. (1999) Curr. Op. Biotechnol. 10:422-427).
Independently of the source of the host cells, exploiting rDNA techniques for the production of recombinant gene products requires a number of considerations. The coding sequence of the gene of interest must be compatible with the chosen host. The starting codon must be unequivocally recognized; codons coding for the amino acids of the heterologous protein must be complemented by anticodons of the host's tRNA and, preferably, should correspond to the most representative tRNAs in the tRNA pool of the said host. Moreover, if present, signals for post-translational modifications must be the same as, or compatible with, those of the host cell. The recombinant gene, having the above characteristics, must be placed under the control of sequence(s) regulating transcription and translation in the host cell. The recombinant gene and the regulating sequence(s) can be referred to as an expression cassette. The promoter and the terminator regulating sequences must be carefully chosen, since they directly affect the expression levels. Other sequences can determine the fate of the recombinant protein, sorting it in the vacuole, into the nucleus, into the mitochondria, or along the secretion pathways. Also, the stability of the mRNA and its affinity for the translation machinery are affected by the nucleotide sequence. Once the expression cassette has been constructed, it has to be inserted into an expression vector and introduced into the recipient host cell. To this aim, several solutions are available.
The expression cassette can directly be inserted in the genome of the host, by means of recombination, which can be either homologous (between two identical sequences, and thus requiring knowledge of the target sequence) or heterologous (at a position in the genome which cannot be controlled). A selective marker is usually required. Auxotrophic markers complement a nutrient request, allowing the growth of the recombinant cells on non-supplemented medium. Dominant markers are genes conferring the resistance to some toxic compounds, so that only cells bearing such a marker can grow on selective media.
Alternatively, it is possible to use episomal vectors, commonly referred to as plasmids. Such vectors are DNA fragments able to replicate themselves in host cells. There are different kinds of episomal vectors, depending on the host. For the well known host S. cerevisiae, exemplary episomal vectors include those reported by Rose, A. B. and Broach, J. R. (1990) Methods Enzymol 85:234-279; Schneider, J. C. and Guarente, L. (1991) Methods Enzymol 194:373-388; Romanos, M. A. et al., supra; and Fukuhara, H. (1995) FEMS Microbiol. Letters 131:1-9. Specific examples include:                In the 2μ-like plasmid, the heterologous gene is inserted in a vector bearing all or part of the sequences from the native 2μ plasmid of S. cerevisiae. The plasmid is extremely stable without selective pressures.        The ARS plasmid is a quite unstable vector based on an ARS endogenous sequence, which promotes DNA replication at the time of chromosomal replication.        The Centromeric plasmid is essentially an ARS vector that is stabilized by the addition of a centromeric sequence (CEN); these CEN sequences drive a correct partition of the vectors during mitosis. It is a very stable vector, but is retained at a low copy number (1-2/cell).        Linear plasmids, which comprise double-stranded DNA or RNA sequences, are quite common in yeasts, and theoretically could be used in strains lacking 2μ-like plasmids. To date, no significant data are available for such a use.        A Minichromosome, based on a very long DNA fragments, is very stable as a result of the presence of telomers, in addition to the above-described ARS and CEN sequences. Minichromosomes are often used for basic research purposes. Potentially they can be used to clone a large cluster of genes, such as a complete heterologous metabolic pathway.        
From the comparison of the expression levels of a large set of recombinant proteins obtained in several host cells, it is apparent that the ideal host cell is not yet available. In fact, each species has some drawbacks that should be carefully evaluated, as each species' drawbacks can be only partially overcome by a good strategy of production. These drawbacks justify research into new host cells, in which negative traits are absent or attenuated.
Bacterial hosts have been used for the production of heterologous proteins, typically the well-known E. coli. Since heterologous products for food and pharmaceutical applications must be free from any toxic or dangerous compounds, bacterial cells do not represent the ideal host for applications in the above-cited industrial sectors. E. coli produces some toxic or potentially toxic metabolites, which must be removed with careful purification protocols. Moreover, yields of heterologous products are often lowered by the formation of large insoluble aggregates, commonly called inclusion bodies. Further, E. coli typically possesses strong proteolytic activity, which can be detrimental to the production of heterologous proteins. S. cerevisiae offers many advantages for the production of recombinant gene products; unfortunately, this host is unable to utilize some very cheap carbon sources, such as starch and whey. Furthermore, this host produces high amounts of ethanol when grown in the presence of relatively high sugar concentrations; the ethanol production (determined by the Crabtree effect) can be overcome with a careful (but not economically feasible) monitoring of the fermentative conditions. The high ethanol production lowers biomass yields and, consequently, yields of the heterologous gene product. Moreover, this yeast has a secretion apparatus not suitable for the very high production levels required for industrial purposes. Finally, the secreted proteins are often hyperglycosylated when compared to the natural product, and therefore it is hard to obtain the production of a protein identical to the original one.
Recently, expression of recombinant gene products has been obtained in some non-conventional yeasts: Hansenula polymorpha, Pichin pastoris, Kluyveromyces lactis and Yarrowia lipolytica (reported in Buckholz, R. G. and Gleeson, M. A. G. (1991) Bio/Technology 9:1067-1072; Fleer, R. (1992) Curr. Op. Biotechnol. 3:486-496; Gellissen, G. and Hollenberg, C. P. (1997) Gene 190: 87-97; Muller, S., Sandal, T., Kamp-Hansen, P. and Dalboge, H. (1998) Yeast 14: 12671283; and Lin Cereghino, G. P. and Cregg, J. P., supra). Said yeasts display some interesting attributes when compared to S. cerevisiae, such as possibly better expression levels, or favorable growth characteristics such as high efficiency of growth on low-cost substrates or ability to grow under severe culture conditions. (Sudbery, P. E., (1994) Yeast 10: 1707-1726, Romanos, M. A. et al., supra; Lin Cereghino, G. P. and Cregg J. P., supra).
Another non-conventional yeast seems to offer many advantages when compared to S. cerevisiae host cells: Zygosaccharomyces bailii. This yeast displays an exceptional resistance to several stresses. For this reason, it is one of the main economically relevant spoilage yeasts. In fact, this yeast can grow in media with low water availability, high hydrostatic pressure (Palou E., Lopez-Malo A., Barbosa-Canovas G. V., Welti-Chanes J., Davidson P. M., Swanson B. G. (1998) J. Food Proto 61:1657-60) and (relative to S. cerevisiae) high temperatures (Makdesi A. K. and Beuchatlo R. (1996) Int. J. Food Microbiol. 33: 169-81). In addition, it tolerates high sugar concentrations. Another remarkable characteristic is its very good tolerance to acidic environments, as it grows at pH values as low as 2, and with high partial CO2 pressures. Further, this yeast can survive to high preservative concentrations, such as 600 mg/l of benzoic acid (Makdesi et al., supra) or to sorbic acid (Cole M. B., Keenan M. H. (1986) Yeast 2:93-100). However, physiology studies and molecular genetic tools for the genus Zygosaccharomyces are very poor: as of Dec. 28, 1999 only 11 genetic sequences from Z. bailii and 38 from Z. rouxii were available at the Internet site GenBank, and only a fraction of the genetic sequences code for proteins. Six different endogenous plasmids have been isolated from the genus Zygosaccharomyces: pSR1, from Z. rouxii, pSB2 from Z. bailii, pSM1 from Z. fermentati, and pSB 1, pSB3 and pSB4 from Z. bisporus. They are structurally and functionally related, but they do not share sequence homology among themselves or with the 2μ endogenous plasmid of S. cerevisiae, so that they usually are not maintained in different species. Plasmid pSR1 (6251 bp), the endogenous plasmid of Z. rouxii, is the most studied in this genus. Its structure resembles that of the 2μ plasmid of S. cerevisiae, and displays a pair of inverted repeat sequences between 2 unique sequences, bearing 3 genes (R, recombinase; P and S, stability) and the sequence Z, a cis-acting locus involved in maintenance of the plasmid. Each of the repeated sequences contains an ARS, which is also recognized by S. cerevisiae (Araki, H., Jearnpipatkul, A., Tatsumi, H., Sakurai, T., Ushio, K., Muta, T. and Oshima, Y. (1985) J. Mol. Biol. 182:191-203). Since the sequences recognized by the recombinase are not completely overlapping (Araki, H. and Oshima, Y. (1989) J. Mol. Biol. 207:757-69), the 2μ plasmid cannot replicate in Z. rouxii. Proteins P, R and S are also characterized in that they cannot complement between Z. rouxii and S. cerevisiae (Araki et al. (1985) supra).
Unlike the plasmid replication origin from S. cerevisiae, the ARS 1 chromosomal replication origin from S. cerevisiae is recognized by Z. rouxii: a centromeric plasmid with this sequence could be stably maintained (Araki, H., Awane, K., Irie, K., Kaisho, Y., Naito, A., and Oshima, Y. (1983) Mol. Gen. Genet. 238: 120-8).
The molecular mechanisms for plasmid replication and repartition are not transferable among the Zygosaccharomyces yeast strains: this is the case of pSR1 and pSB3 (Utatsu, I., Utsunomiya, A., and Toh-e, A. (1986) J. Gen. Microbiol. 132:1359-65). This fact is not surprising, as sequences of all the isolated plasmids are very different, except for a certain homology between pSB1 and pSB4. Usually, S. cerevisiae is the less restrictive yeast, as it can recognize ARS sequences from other sources; on the other hand, no one of the Zygosaccharomyces strains tested so far was able to replicate a plasmid bearing ARS sequences from the S. cerevisiae 2μ natural plasmid.
Few attempts were made to apply knowledge of the genetics of the yeast Z. rouxii for the production of heterologous proteins. The only example concerns the expression of alkaline protease from Aspergillus oryzae. The expression system is based upon the endogenous pSR1 plasmid, and the use of the endogenous promoter GAPDH; geneticin (G418) resistance is the dominant marker (Ogawa, Y., Tatsumi, H., Murakami, S., Ishida, Y., Murakami, K., Masaki, A., Kawabe, H., Arimura, H., Nakano, E., Motai, H. et al., (1990) Agric. & Biol. Chem. 54:2521-9).
Knowledge of the genetics of Z. bailii is even lower. The endogenous plasmid pSB2 (5415 bp) shows some analogies with pSR1 (Toh-e, A., Araki, H., Utatsu, I., and Oshima, Y. (1984) J. Gen. Microbiol. 130: 2527-34; Utatsu, I., Sakamoto, S., Imura, T., and Toh-e, A. (1987) J. Bacteriol. 169:5537-45). In addition, some linear double stranded RNA plasmids have been described. (Radler, F., Herzberger, S., Schonig, I., and Schwarz, P. (1993) J. Gen. Microbiol. 139:495-500). Very recently, some information appeared about the development of genetic tools for Z. bailii. A genomic bank of the yeast has been obtained (Rodrigues, F., Zeeman, A. M., Sousa, M. I., Steensma, H. Y., Corte-Real, M., and Leao, C. (1999) In: Proceedings of the XIX International Conference on Yeast Genetic and Molecular Biology, Curr. Genet. 35:462) and the disruption of the URA3 gene has been also described (Mollapour, M. and Piper, P. W. (1999) In: Proceedings of the XIX International Conference on Yeast Genetic and Molecular Biology, Curro Genet. 35:452). However, nothing has to date been published about homologous or heterologous protein expression obtained from this yeast.
Since Z. bailii yeasts can grow in very restrictive cultural conditions (e.g., pH, ionic strength, temperature, sugar concentration, and acid concentration), they are potentially interesting from an industrial point of view. In fact, the features described above greatly simplify many fermentation procedures. For example, there is no need for strict and sophisticated control of process parameters and of medium composition. Moreover, the ability to grow at higher temperature facilitates heat control, one of the primary problems arising during high density, large-scale fermentations. Finally, fermentation in restrictive conditions prevents contamination problems, thus reducing the need for expensive sterilization steps. All those elements, together with a high specific productivity, are essential to allow the economic success of an industrial production of heterologous protein.
In view of the above considerations, the importance of the development of genetic expression system(s) for Zygosaccharomyces bailii strains and a fast and reliable transformation protocol for the production of recombinant proteins (i.e., heterologous and/or homologous) in such a host are clear.